AVS2010 Session AS-WeM: New Ion Beam Technologies for Imaging, Sample Preparation and Analysis

The development of a new radiation effects microscopy (REM) technique is crucial as emerging semiconductor technologies demonstrate smaller feature sizes and thicker back end of line (BEOL) layers. To penetrate these materials and still deposit sufficient energy into the device to induce single event effects, high energy heavy ions are required. Ion photon emission microscopy (IPEM) is a new technique that utilizes coincident photons, which are emitted from the location of each ion impact to map out regions of radiation sensitivity in integrated circuits and devices, circumventing the obstacle of focusing high-energy heavy ions. The (x,y) coordinates are instead determined with a single photon, position-sensitive detector. Thus, a high energy broad beam can be used to achieve high LETs, while still mapping out radiation-sensitive regions with sufficient resolution. Several versions of the IPEM have been developed and implemented at Sandia National Laboratories (SNL). The initial IPEM was a tabletop system, which utilized a Po-210 alpha source as the incident radiation. The second version has been utilized on the microbeam line of the 6 MV tandem accelerator at SNL, which allows for direct results comparisons between data obtained with a scanned, focused microbeam, and that from IPEM. Another IPEM was designed for ex-vacu use at the 88” cyclotron at Lawrence Berkeley National Laboratory (LBNL). That facility allows for the use of a heavy ion cocktail, with beam energies up to several GeV. Extensive engineering is involved in the development of these IPEM systems, including resolving issues with electronics, event timing, optics, phosphor selection, and mechanics. The various versions of the IPEM and the obstacles, as well as benefits associated with each will be presented. In addition, the current stage of IPEM development as a user instrument will be discussed in the context of recent results.

*Sandia National Laboratories is a multi-program laboratory operated by Sandia Corporation, a wholly-owned subsidiary of Lockheed Martin company, for the U.S. Department of Energy's National Nuclear Security Administration under contract DE-AC04-94AL85000.

Milling speeds with a gallium focused ion beam (FIB) are often much too slow for many sample preparation and surface engineering applications. For example, cross-sectioning stacked-die semiconductor devices, prototyping micro-mechanical structures and delayering IC’s for circuit mapping are growing applications that require a milling rate that far exceeds that provided by the gallium FIB.

Furthermore, secondary ion mass spectrometry (SIMS) imaging has been limited to a lateral resolution of 200nm when using an oxygen focused ion beam for high sensitivity surface analysis. Many applications in material science could benefit from an ability to image trace level surface chemistry with <20nm resolution. Example applications include, sub-cellular imaging of trace metals in the brain for neurodegenerative disease studies, analysis of trace element segregation in metal alloys and studying isotope distributions in meteorites and interplanetary dust particles.

In the more general area of direct-write surface engineering, precision milling and deposition with nanometer precision is limited to volumes of <10⁴um³ when using gallium FIB systems. Also, engineered devices must generally be tolerant of high gallium concentrations being implanted in the near-surface region. These are major restrictions when fabricating devices that require nanometer precision across dimensions of several hundred micrometers. Additionally, inherent gallium implantation can render structures bio-incompatible and compromise the electrical properties of many materials.

Here, we review a plasma ion source technology (Hyperion^TM) that can provide a focused ion beam capable of milling silicon at a rate of >5000um³/s with <4um milling resolution and <20nm imaging resolution with 30keV xenon ions. The same ion source is also readily operated as a high brightness source of oxygen, hydrogen and any inert ions.

By transferring energy to plasma electrons via a RF induction field, it is possible to create a plasma state without a cathodic electrode. This approach can create high plasma densities (>1x10¹³ ion cm^-3), with very low mean thermal ion energies (<0.05eV), providing the conditions for an energy normalized beam brightness that now exceeds 1x10⁴ Am^-2sr^-1V^-1. This high brightness can be attained with long lifetimes (>>2000 hours), stable beam current (<±0.5% drift per 30 minutes) and an axial energy spread for the extracted beam of 5-6eV and for an array of ion species.

This paper presents FIB and SIMS data from this new ion source. The operating principles of the ion source, the properties of the beam(s) being created and the projected future for this technology are also described.

When sputter profiling a multi-layer material using XPS, it is often difficult to obtain an accurate calibration of the depth scale. In part, this is due to the fact that the sputter rate in each material is different and a post profile measurement of crater depth will only yield an average value for the sputter rate. The purpose of this paper is to provide a method for an internal, standardless calibration of the depth scale which can be applied to single profiles through multilayer materials.

As a layer is etched to a thickness of a few nanometres, the spectrum from the underlying layer can be seen. With knowledge of the electron attenuation lengths in the layers it is then possible to calculate the thickness of the remaining upper layer. If this calculation is carried out following each of a number of sputter cycles then the sputter rate within the layer can be calculated. This value can then be applied to sputtering within the whole of the upper layer. If this is repeated for each layer in a multilayer sample then an accurate depth scale can be constructed.

This method will be applied to a number of samples including standard multilayer materials and the value of the method assessed.